ADVERTISEMENT

U.S. Will Join G-7 AI Pact, Citing Threat From China

U.S. Will Join G-7 AI Pact, Citing Threat From China

(Bloomberg) -- The U.S. plans to join an international organization that will advise companies and governments on the responsible development of artificial intelligence, becoming the last of the Group of Seven countries to sign on.

The group, called the Global Partnership on AI, will study and provide recommendations to encourage the creation of AI technologies that respect privacy and civil liberties, said Michael Kratsios, the chief technology officer of the U.S. He called the partnership “a first of its kind global organization that sees the future of AI as something that can uplift Americans and people around the world.”

U.S. officials plan to outline their involvement in the program on Thursday at a G-7 meeting of science and technology ministers. In addition to G-7 members, several more countries outside the group, including New Zealand, are expected to eventually join the program. Kratsios acknowledged that the Trump administration had generally been skeptical of multilateral agreements but said that this one was important as a check on China’s approach to AI, which often involves using new technologies to augment an already-robust surveillance state. The Chinese government, Kratsios, said, “has twisted AI in ways that are in direct conflict with the values of the U.S. and its allies.”

Tensions between the U.S. and China, heightened throughout Donald Trump’s presidency, were further escalated in the last week over a Chinese national security law that tightens Beijing’s grip over Hong Kong. Michael Pompeo, the U.S. secretary of state, said Wednesday that the country could no longer certify Hong Kong as politically autonomous, a move that could trigger sanctions.

Kratsios compared the new AI partnership to past U.S.-led efforts to contain China’s influence over 5G wireless technology. Last year, the U.S. and 31 other countries met in Prague and agreed to a set of cybersecurity standards that included warnings about suppliers from countries that do not have adhere to international standards on security and data protection. This was widely seen as a reference to Huawei Technologies Co. and ZTE Corp., the big Chinese telecom equipment makers. The U.S. has tried to keep the companies out of its own country and international markets, fearing they would create backdoors for Beijing. Both companies have vigorously denied the charges.

The idea of both the 5G and AI initiatives is to set limits on Chinese companies seeking to expand abroad, said Kratsios, a former chief of staff at Thiel Capital, an investment fund backed by Peter Thiel. For years, American tech executives have suggested that Chinese influence could erode free speech around the world because companies in China are required by law to censor politically sensitive topics while allowing authorities access to user data.

The White House has expressed similar concerns. In a speech last year, Kratsios warned that adoption of Chinese standards in 5G and AI could “run the risk of repeating the same mistakes our nations made nearly 20 years ago,” when China joined the World Trade Organization.

The G-7 meeting on Thursday will be led by Kratsios and Kelvin Droegemeier, Trump’s science adviser. The new group will ask member states to follow principles drafted by the Organization for Economic Cooperation and Development, which recommends developing AI technologies that respect human rights and are transparent to anyone affected by them. The U.S. had initially refused to join the new G-7 partnership, it said, in part out of a concern that the group would conflict with the OECD’s framework. The group’s members will be nominated by each country and composed of experts from academia, business and nonprofits.

Ahead of the meeting, Kratisos said that Covid-19 had made fears about China more urgent, pointing to reports that Beijing had been silencing critics of its coronavirus response, as well as a new software system that tracks citizens and automatically assigns them a color-coded health score, which is then used to limit their movements. “This Covid crisis has put in focus a stark choice,” Kratsios said. “AI must not be a tool to repress people and invade their privacy.”

©2020 Bloomberg L.P.